MEGA-Bench Task - Red Teaming Captcha — leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.0 Flash (Preview) | 13.2 |
| 2 | Qwen 2 VL 2B | 12.1 |
| 3 | InternVL2.5-2B | 11.1 |
| 4 | Gemini 2.5 Pro | 10 |
| 5 | GPT-4o | 10 |
| 6 | Gemma 3 27B (IT) | 10 |
| 7 | Gemma 3 12B (IT) | 10 |
| 8 | Gemma 3 4B (IT) | 10 |
| 9 | InternVL3-78B | 10 |
| 10 | InternVL2-8B | 10 |
| 11 | InternVL3-8B | 10 |
| 12 | Qwen 2 VL 72B | 10 |
| 13 | Qwen 2 VL 7B | 10 |
| 14 | InternVL3-38B | 10 |
| 15 | InternVL3-14B | 10 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-red-teaming-captcha · How the rankings work · Data refreshed daily, snapshot 2026-07-22.